Feng Liu (he/him) @ The University of Melbourne

Home
Research Focus
Research Group
Teaching
Publications
Professional Services
Miscellaneous
Join Us

Home


Feng Liu

Feng Liu, Ph.D.

Statistically grounded trustworthy machine learning for modern AI systems.

I lead research on rigorous methods for evaluating, adapting, and safeguarding AI systems under distribution shift, privacy risk, and real-world uncertainty.

Senior Lecturer in Machine Learning and ARC DECRA Fellow,
School of Computing and Information Systems, The University of Melbourne
Co-Director, Trustworthy Machine Learning and Reasoning (TMLR) Lab

Visiting Scientist @ Imperfect Information Learning Team, RIKEN-AIP
Visiting Fellow @ DeSI Lab, Australian Artificial Intelligence Institute, UTS

Room 3317, Level 3, Melbourne Connect (Building 290), 700 Swanston Street, Parkville VIC 3010, Australia.
Academic: fengliu.ml [at] gmail.com | feng.liu1 [at] unimelb.edu.au
Industry: fengliu.genai [at] gmail.com
[Google Scholar] [GitHub] [Lab]

Research Vision

I build statistically grounded trustworthy AI. My research develops rigorous methods for evaluating, adapting, and safeguarding modern AI systems under distribution shift, privacy risk, and real-world uncertainty.

My current program focuses on data-adaptive hypothesis testing as a foundation for machine learning, and certified evaluation of large language models with statistical guarantees.

Current Primary Priorities

  • Data-adaptive hypothesis testing for machine learning. I study adaptive testing procedures for modern ML systems, including e-processes for sequential and streaming settings and learned metrics / learned representations for measuring distributional closeness, dependence, and reliability.

  • Certified evaluation of LLMs with statistical guarantees. I develop statistically principled methods for evaluating safety, jailbreak robustness, privacy leakage, unlearning, and deployment-time reliability of foundation models.

These priorities sit within a broader research program spanning trustworthy AI, statistical machine learning, and reliable deployment. A fuller overview is available on the Research Focus page.

Selected Leadership & Impact

  • Research leadership: Co-Director, Trustworthy Machine Learning and Reasoning (TMLR) Lab; Program Co-Chair, AJCAI 2026; Communication Chair, NeurIPS 2026; Co-Chair, ICML 2026 Workshop on Hypothesis Testing.

  • Editorial roles: Action Editor, Transactions on Machine Learning Research; Action Editor, Neural Networks; Editorial Board Member, Machine Learning; Editor, ACM Transactions on Probabilistic Machine Learning.

  • Selected recognition: NeurIPS 2022 Outstanding Paper Award; Top Area Chair Award, NeurIPS 2025; Outstanding Area Chair Award, ACM MM 2024; ARC DECRA Fellow; FEIT Excellence Award in Research.

Collaborations

I welcome collaborations on trustworthy AI, statistical machine learning, data-adaptive hypothesis testing, certified LLM evaluation, and the responsible deployment of AI in science, education, and other high-stakes domains.

Prospective students and collaborators are welcome to visit the Join Us page for future opportunities and project-based collaboration.

Short Biography

I am a Senior Lecturer in Machine Learning and ARC DECRA Fellow at the School of Computing and Information Systems, The University of Melbourne, where I co-direct the Trustworthy Machine Learning and Reasoning (TMLR) Lab. My research develops rigorous statistical foundations and practical methods for trustworthy AI, with representative publications in Nature Communications, Nature Plants, JMLR, TPAMI, TNNLS, NeurIPS, ICML, and ICLR.

Research Milestones

Research Experience

Education

  • Ph.D. in Computer Science (November 2020)

    • Faculty of Engineering and Information Technology,
    University of Technology Sydney, Sydney, Australia.
    Supervised by Dist. Prof. Jie Lu and Prof. Guangquan Zhang

Sponsors

Australian Research Council CSIRO NSF